There is an abundance of literature dating back over the past 20+ years about the challenge of premature gearbox failures, and the cost impact they have on wind turbine operation. While the principals of prognostics and health management (PHM) are well established, and the objective of replacing unplanned failure events with scheduled maintenance based on early indication of degradation has not changed, the wind industry and sensor technology continue to evolve in ways that steadily increase the value proposition.

With global acceptance of the need to shift our energy dependence to renewables, the demand for wind energy is driving the development of much larger turbines and a significant increase in offshore wind farms. The primary cost avoidance targets associated with PHM, or condition based maintenance (CBM) are tied to business interruptions, inspection and repair costs, along with downtime penalties. The larger the turbine, the more difficult to reach, the higher the cost and complexity associated with inspection and maintenance. Secondary or catastrophic failure events that cannot be addressed in-situ are even more concerning with taller, harder to reach and heavier components. Further, with greater dependency on wind as a primary energy source, the cost of downtime penalties is also likely to continue increasing.

Evolution of wind turbine size and output. Source: Liebreich

Wind turbine heights and rotor diameters have easily doubled since the early 2000s as the industry pushes the boundaries of production per turbine. With the emergence of offshore wind as a major source of energy, size will continue to increase the maintenance challenge. In 2019, General Electric installed the Haliade-X turbine prototype in the Port of Rotterdam. The wind turbine stands 260 m (853 ft) tall and has a rotor diameter of 220 m (721 ft). Vestas plans to install the V236-15MW offshore prototype at Østerild National test centre for large wind turbines in Western Jutland, Denmark, in the second half of 2022. The wind turbine is 280 m (918 ft) tall with a predicted production output of 80 GWh a year, enough to power nearly 20,000 households. The criticality of keeping the massive wind turbines of the future running and being able to efficiently maintain them based on condition will only increase in importance.

While wind turbine, gearbox and bearing manufacturers strive to design more reliable assets, the common phenomenon limiting useful life continues to be surface fatigue resulting from repeated stresses under bearing rolling contact or gear meshing contact. Excessive loads, misalignment, material flaws, manufacturing defects, mishandling, contaminants in the oil, high oil temperatures and corrosion are some of the potential contributors to localized damage that begin to degrade the bearing or gear. In other words, the reality is that even with more reliable bearings and gearboxes, there will always be a probability of failure over time supporting the value proposition of moving from reactive to proactive, condition-based and ultimately to predictive maintenance.

An offshore wind installation near Scotland

That desired future state of truly predictive maintenance is well articulated in the principal of “Maintenance 4.0” that is the application of “Industry 4.0” technologies (industrial analytics, automation, robotics, etc.) to operations and maintenance (O&M) activities. The business objective is to radically improve equipment availability while lowering O&M costs through digitalization. While much attention is paid to the advances in artificial intelligence as means to turn data into valuable insights, the often unspoken challenge is that insights are only as good as the data.

In the case of wind turbines, that data comes from sensors used to monitor the health of the bearings and gearboxes. Progressing toward predictive maintenance requires the ability to not just identify damage, but to determine damage severity and calculate time to failure or remaining useful life (RUL). This is where oil debris monitoring differentiates itself from other types of sensors, as quantifying the wear debris from a damaged component is a direct representation of the damage of the monitored component.

For wind turbine gearboxes, oil debris monitoring (ODM) technology provides an early indication of bearing spall and gear pitting damage and quantifies the severity of damage progression towards failure. Online oil debris monitoring provides the most reliable and timely indication of bearing degradation because:

• Bearing failures on rotating machines tend to occur as events and could be missed by means of only periodic inspections or data sampling observations.
• As large wear particles are being detected by the oil debris monitoring sensor, there is a low probability of false indication from benign smaller wear debris particles.
• Residual or wear-in debris can be differentiated from the actual damage debris because the accumulated particle counts recorded due to the former tend to decrease while those due to the latter tend to increase.

While ODM technology itself is a proven technology used in asset management across multiple industries, truly understanding the physics of failure to be able to develop the algorithms necessary to accurately predict RUL requires years of study. Ottawa, Canada-based Gastops first started conducting surface fatigue tests on bearings to understand the progression of failure in 1992, bringing ODM to the wind market in the early 2000s.

The company is now a world leader in critical equipment condition monitoring, having installed over 20,000 oil debris monitoring sensors on active wind turbines with a proven history of detecting early signs of damage in advance of failures, maximizing availability while minimizing downtime and maintenance costs. Realizing the goals of PHM, the technology is used to monitor drivetrain health, by detecting the initiation of damage and monitoring its progression, which enables maintenance events to be scheduled proactively, preventing costly unplanned downtime during critical operating periods.

The ability to determine the remaining useful life of an asset is critically important as it provides the necessary information for operators to optimize the operational lives of their gearbox and main bearing. The company’s flagship product, MetalSCAN, is an online advanced oil debris sensing technology which is plumbed into the return oil line from the gearbox or main bearing. Any debris generated within the gearbox passes through its core and is quantified in terms of size, frequency and the type of metal, whether it is ferrous or non-ferrous material.

This information is collected and compared against predefined warning and alarm limits, developed using customized algorithms based on the geometry of the gearbox. When debris passes through the sensor, MetalSCAN translates the information it gathers into the RUL of the gearbox or main bearing. This allows the operator to understand the current health, adjust operational parameters and plan maintenance activities, preventing the requirement for shutdown during critical operation periods.

While ODM provides an ideal data source to provide the valuable insights required to support predictive analytics, realizing the vision, and leveraging the potential of Maintenance 4.0 will require a variety of data sources in multiple locations to improve the accuracy of modeling, allowing operators to pinpoint failures throughout the system. Many wind turbine manufacturers already combine vibration sensors and ODM into the overall condition monitoring systems they provide.

As technology advances to enable the transition from condition-based maintenance to truly predictive maintenance, real-time equipment intelligence will be required. Systems that combine the information from next generation real-time sensors that monitor data on multiple factors such as oil debris, oil condition, vibration, temperature and pressure will be used to develop advanced analytics and digital twins. Connectivity to enable large scale IIOT deployments with the assurance of data security and network reliability will be key factors in enabling wind farm operators of the future to realize the vision of Maintenance 4.0.

Paramount to success in this evolution remains the intersection of machine intelligence with human ingenuity. The expertise that comes with decades of research into the physics of failure by a company like Gastops combined with a commitment to innovation and drive to realize the vision of real-time prognostics will lead to a future of optimized O&M costs in the wind industry. Today’s ODM technology is already providing the benefits of PHM envisioned 20 years ago. The rapidly increasing demand for wind energy is driving the creation of massive wind turbines that are being located offshore. There is a clear requirement to evolve condition monitoring systems to provide operators with PHM capability enabling the future of predictive maintenance.

Jordan Freed is the Director of Corporate Marketing & Product Strategy for Gastops Ltd, where he is responsible for product management, strategic growth and marketing.  With a Bachelor’s Degree in Electrical Engineering and over 25-years of experience, Jordan is a customer focused leader, passionate about innovation and product realization to address the future needs of the market.   His career spanned multiple industries including steel, telecom and defense prior to joining Gastops in 2020 to support the organization’s long term growth objectives.

The post Oil debris monitoring saves time when it comes to wind turbine gearbox maintenance appeared first on Windpower Engineering & Development.

In 2019, Croda acquired Rewitec GmbH and began to offer energy technologies customers nano- and micro-particle-based additives to increase the durability of machinery by lowering friction and reducing wear. Sentient Science validated Rewitec’s DuraGear gearbox oil additives for use in wind turbine gearboxes in 2017.

This partnership will see Sentient apply physics and data science expertise, combined with Croda’s Rewitec additives, to calculate the lifetime extension of critical rotating components. It will also examine how Rewitec’s GR400 grease additive, developed specifically for main bearing durability improvements, can improve equipment lifetime.

Sentient Science provides DigitalClone for wind O&M, which uses a unique combination of physics and data science to give a holistic view of the health and remaining useful life of an asset’s critical systems and components. This information is used for predictive maintenance programs to reduce operations and maintenance costs and ultimately to prolong asset life. Sentient is able to calculate and demonstrate durability improvements imparted through using Croda’s Rewitec technology, which provides asset owners the option of extending the lifetime of their assets instead of costly part replacements.

“The competitive energy market is forcing energy producers to optimize maintenance practices and reduce operational expenses,” said Scott Gardiner, Business Development Specialist, Energy Technologies at Croda. “Major correctives are the largest cost drivers in the wind energy market, specifically gearbox or main shaft replacement.  The cost of this replacement can completely change the asset’s profitability during its lifetime. The Rewitec technology is currently helping customers reduce failure rates and extend the life of these critical assets.  We are excited that customers can now utilize Sentient’s DigitalClone to provide RUL projections in conjunction with our Rewitec technology.”

“As wind turbines age, operators are seeing a higher number of onshore and offshore wind assets running with damage, specifically in critical rotating components like gearboxes and main bearings,” said Ed Wagner, GM of wind operations at Sentient. “Our customers have been waiting for data to compare next generation additives, like Rewitec, against uptower part replacements. And while this may not be a solution for every wind turbine, we do have data to substantiate improvements in gearbox life and expect to show the same in main bearing life.”

News item from Croda

The post Sentient Science and Croda partner for wind turbine life extension appeared first on Windpower Engineering & Development.

Wind turbine gearboxes designed to last two decades are showing damage often within five to 10 years, leading owners and operators to seek stronger solutions.

The evolution of wind turbine gearbox design has resulted in the use of an increased number of tapered roller bearings throughout the parallel shaft section to effectively manage both the radial and axial loads produced by helical gearing during operation. This arrangement includes a cylindrical roller bearing in the float position coupled with a two-row tapered roller bearing in the fixed position, typically in a direct-mount configuration (also referred to as a face-to-face, DF or X arrangement). Compared to conventional arrangements that commonly use various cylindrical, spherical, ball and thrust roller bearing configurations, this stronger approach can better support varying loads experienced during gearbox operation, helping wind farm owners and operators avoid costly repairs and downtime.

For service technicians and rebuilders, this trend toward two-row tapered roller bearings requires that close attention is paid to controlling the assembly’s associated mounted setting. Unlike many conventional bearing types, two-row tapered roller bearing designs are composed of separable components with an unlimited range of axial locations relative to one another based on the installation method used.

A floating cylindrical roller bearing (left) paired with a fixed two-row tapered roller bearing (right) is becoming the preferred method for managing loading conditions seen in wind turbine gearboxes.

These separable components require an external means of controlling the final relative axial location and resulting mounted setting, which is most commonly done by introducing a component spacer into the assembly. This spacer is then specifically machined to a controlled width by the installer or by the bearing supplier during the assembly production. However, many MRO professionals with limited tapered roller bearing experience have questions about how to achieve the best mounted setting results. And without proper mounted setting control, the resulting stresses can become excessive or may even result in unloaded conditions inside the bearing, leading to roller and cage damage.

This article reviews three common approaches for obtaining a final axial setting for a two-row tapered roller bearing assembly using a controlled spacer width.

Roller angles, among other variables, can be customized, allowing two-row tapered roller bearings to better handle combined loads compared to traditional cylindrical, spherical and ball bearing designs.

Green spacer assemblies

A green (or unground) spacer assembly is intentionally supplied with a spacer that has extra width to allow the user to tailor various settings based on specific application requirements. Note that in some instances, bearings may be supplied without a spacer (typically when the user intends to produce or purchase one separately), in which case the combination of bearing components and spacer would be treated similarly when obtaining the target nominal mounted setting. Two common methods of obtaining the target mounted setting for a green spacer assembly are the measurement-and-calculation method and the manual push-pull method.

Measurement-and-calculation method

The measurement-and-calculation method requires the user to obtain measurements of the shaft, housing, bearing bore and bearing outside diameter (O.D.) to determine the actual mating component fitting practice. Secondly, the user must obtain the correct lateral loss factor(s) for the bearing assembly part number to support an accurate lateral loss calculation due to any resulting interference fits. Finally, the spacer gap between the bearing components at zero bench setting must be calculated using assembly component physical measurements (also known as drop measurements) or measurements obtained from the bearing supplier, thus allowing the user to determine the final spacer width to reach the target nominal mounted setting.

The manual push-pull method measures total shaft axial movement using a dial indicator.

Push-pull method

Alternatively, the manual push-pull method requires the user to mount the bearing components and the unground spacer using the actual mating shaft and housing components to determine a baseline mounted setting value. This baseline setting can then be compared to the target nominal mounted setting to determine the spacer width that must be ground to achieve the target setting.

The push-pull method requires all rollers to be seated completely and uniformly against the adjacent large rib and requires that no external components or loads interfere with the ability to obtain an accurate measurement of the axial shaft movement in both the push and pull steps of the process. To ensure that rollers are seated and an accurate axial setting is measured, an axial load must be applied to the shaft in one direction while the shaft is oscillated a minimum of 20 times before applying an axial load to the shaft in the opposite direction while oscillating the shaft as before. The total axial movement of the shaft is the bearing assembly mounted setting.

Assuming proper procedures, a green spacer makes it possible to achieve an optimal final mounted setting, given that the actual dimensions from the mating bearing components, shaft and housing are accounted for, thus eliminating (most) tolerances from the equation and resulting in a very small possible mounted setting range. However, inaccurate measurements, incorrect calculations, imprecise spacer grinding, mismatched mating components or the inability to obtain an accurate push-pull measurement due to component size or adjacent component interference will directly impact the ability to achieve a high level of precision and consistency, leading to an unpredictable possible mounted setting range that can lead to a range of problems.

Matched spacer assemblies

A matched spacer assembly (also referred to as a single bench setting assembly) is supplied with a fixed width spacer that has been ground by the bearing supplier to a specific width based on an assumed set of component fitting practice values and a target nominal mounted setting value for the application.

For this type of assembly, the bearing production facility completes the bearing physical measurements to understand the spacer gap between components at zero bench setting and to determine the final spacer width necessary to reach the specified nominal mounted setting. Both the assumed fitting practice range and the calculated lateral loss would be considered by the bearing supplier to determine the bench setting for the matched bearing assembly.

A matched spacer could be considered the least accurate approach to obtaining a target nominal mounted setting given that the shaft, housing, bearing bore and bearing O.D. tolerances all have a direct effect on the possible final mounted setting and are not accounted for on an individual combination of mating components, thus resulting in a relatively large possible mounted setting range.

Bore-compensated spacer assemblies

A bore-compensated assembly (also referred to as a variable bench setting assembly) is supplied with a fixed-width spacer that has been preground to a specific width based on an assumed set of component-fitting practice values and a target nominal mounted setting for the application.

To begin, the bearing supplier again completes drop measurements to determine the spacer gap between the components at zero bench setting and final spacer width to reach the target nominal mounted setting. The difference between this assembly type and a matched spacer, however, is that the potential interference fit range between the shaft and inner ring bore is broken down into multiple groups, such that a different bench setting value can be selected based on the actual component bore dimension (rather than using a single bench setting for the entire interference fit range). This results in a more tightly controlled selection of the final spacer width and a reduced target mounted setting range.

Typically, the target mounted setting and shaft tolerance are obtained from the user or from application engineering review. Engineering then performs calculations that cover the known inner ring fitting practice range to develop a table of bench setting values. This table allows the bearing supplier to select a single bench setting based on the actual bearing bore measurement and associated shaft tolerance combination. The bearing supplier then completes the measurement process to understand the dimension of the spacer gap at zero bench setting to thus determine the final spacer width necessary to achieve the target nominal mounted setting.

Bore-compensated assemblies are considered more accurate than matched assemblies, given that the bore tolerance variable and associated lateral loss are already accounted for in the fitting practice; however, this method cannot achieve the extreme accuracy of a properly controlled and processed green spacer assembly.

Note that this assembly type typically assumes a loose fit or a very light interference fit between the bearing outer ring and housing. In other cases, special compensation matrices may be necessary to accommodate interference fits at both the bore and O.D. of the bearing assembly.

The following table can help you understand the relative advantages of different assembly types (on a four-point scale, one star is acceptable and four starss is optimal):

Avoid these costly mistakes

Bearings that use a spacer to control the final mounted setting may be unfamiliar to many service techs. Keep the following advice in mind when maintaining or upgrading gearboxes:

• Do not use an old spacer in a new bearing assembly.
• Do not swap spacers between old and new assemblies.
• Do not install an assembly without a spacer unless another axial setting mechanism is present (e.g., spring-loaded system or component end-cap).
• Never install a green spacer that has not been ground to proper width.
• Never assume a supplied spacer is the correct finished width.
• Always take time to measure and verify.

Where to turn

It is wise to consult a qualified expert to review the described measurement, calculation and assembly procedures when installing bearings that rely on a finished spacer width to control the mounted setting. Your trusted adviser can provide the support and clarity needed to ensure a successful outcome that avoids major corrective work. As the use of two-row tapered roller bearing assemblies in wind turbine gearboxes becomes more prevalent, it is only a matter of time until such questions arise. Be sure to ask your bearing supplier about training and education that can put your team ahead of the curve.

Kyle Smith is a principal application engineer for The Timken Company with a current focus on wind gearbox applications solutions and upgrades. He is based out of the company’s world headquarters in North Canton, Ohio. Kyle has held past engineering roles in support of various industrial markets, including power transmission equipment, aggregate and crushing equipment, paper production equipment and heavy moveable structures. He has been with Timken for 14 years. Kyle earned an associate degree in mechanical engineering technology from Stark State College and a bachelor’s degree in manufacturing engineering technology from the University of Akron.

The post What to know when setting two-row tapered roller bearings in wind turbine gearboxes appeared first on Windpower Engineering & Development.